Four Gold Medals and a Forgotten Data Column: Vietnamese Athletics Through the SEA Games Lens

Phnom Penh, May 9, 2026, early afternoon. Nguyen Thi Oanh had just won the 1500m final. A few hours later she stepped into the 3000m steeplechase final and won again. No big screen on the grandstand replayed that closing surge. In the press room, reporters asked how she felt. Nobody asked whether the track was dry or damp, whether the wind was against her or behind her, what threshold her heart rate hit on the final lap. The results sheet carried a single line: first. That night, from Hai Phong, I opened my four-SEA Games data file and started counting strides again. I came to athletics from a football pitch, not from a running track. In 2026, while working as a data consultant for the youth side of a Hai Phong football club, I found a pressing midfielder with an average PPDA of 6.8 — the best in the academy — who drew almost no attention because of his modest frame. I carried the spreadsheet into the meeting room and asked for him to be given a chance. In the match against Hanoi FC on V.League round 17, he won the ball 14 times, provided one assist, and the team won 2-1. Since that day I have understood one thing: crowds see the goal, data sees the entire 90 minutes of pressure behind it. When the pandemic stopped football in 2026, I lost my live data feed. I spent four months reviewing 2,300 matches from five domestic seasons and three major European leagues, building a pressure index that combined PPDA, defensive distance and pressing speed. Teams with a PPDA below 8.5 averaged 1.8 points per match, well above the rest of the dataset. But when I moved to athletics, I had to start learning again. A football match has 22 players and thousands of events to measure. A running lane has one person and one number. The paradox is this: the fewer the variables, the easier it is to misread. That is why I built my own athletics database. No provider in Vietnam sells data packages on final-lap speed or steeplechase rhythm from the SEA Games. I had to extract it myself from official results, from video recordings, from organisers' releases. For each SEA Games edition I recorded four columns: gold medals by event; the number of different athletes who won any medal; the number of events in which the delegation placed at least two athletes in the final; and the number of personal bests set at the Games themselves. Together these four columns make what I call the depth index. It does not replace the medal table. It answers a question the medal table does not: is the system strong at the peak, or strong at the base. The last four SEA Games, from Singapore 2026 to Phnom Penh 2026, give me a far clearer picture than the gold table in the newspapers. Vietnamese athletics won more than ten gold medals in Phnom Penh, an impressive number. But split by event, most of those golds sat with a very small group of athletes, concentrated in middle distance, steeplechase and a few technical events. Counting the faces that reached the podium, Vietnam ranked below Thailand, even though its gold count was higher. That is the familiar paradox of a system with a sharp peak and a thin base. In women's middle distance, one athlete carries several events and faces almost no regional opposition. In the sprints, the delegation has a few names, but its best personal marks have tended to stall around a certain threshold across several seasons, while average lap speed shows no leap. In throws and jumps, the number of final appearances is fairly stable, but the number of personal bests set at the Games themselves is low. In other words: the peak holds, while the base does not thicken to match. That leads to a question I always put to any dataset: is the number reflecting ability, or reflecting the structure of the competition. The SEA Games is a stage with thin opponent density in many events. An athlete good enough at 1500m can enter three events and take home three golds. In a denser field those golds would be redistributed. Gold medals measure an individual's absolute ability. Headcounts measure a system's ability. I care more about the second column, because it forecasts the next four years far better. Set against Thailand and the Philippines, the gap becomes clearer. Thailand produces more individual athletics medallists at a single SEA Games than Vietnam, spread evenly across throws, jumps, sprints and distance events. The Philippines has an edge in speed and jumping events, with a younger pool. Vietnam concentrates its strength in middle distance, where a few outstanding individuals pull the whole index up. This is a pattern I meet constantly in football: a team can boast 60% possession built on meaningless sideways passes. Possession is the most deceptive metric I have ever worked with, and an athletics medal table shares one thing with it — both look good on the surface. The men's side gives a similar signal. The number of men's finals is stable, but the count of marks meeting Asian Games or World Championships standards remains small: I can count only a handful across a whole four-year cycle. That says nothing about the athletes' talent. It says the competitive threshold at continental level sits far above the regional one, and a SEA Games gold does not automatically convert into a ticket to a major championship. The combined events — the 400m hurdles and the heptathlon — are where I see the clearest untapped potential, and where few people look. These demand technique, endurance and speed at once, and they are usually overshadowed by the more emotional middle-distance races. Across the last four SEA Games, the number of Vietnamese athletes ranked in this group has not grown, but the average mark has. A small number rising quietly is the kind of signal I value most. I understand why fans remember only the golds. They are right to. A gold medal is a real moment, measurable, undeniable. I am not arguing with that joy. I am only saying that the gold table and system health are two different layers of data, just as the scoreline and actual performance are two different layers. In 2026, before the World Cup, I published an analysis based on Germany's qualifying pressing numbers: an average PPDA of 9.2, far too high for a champion's standard, slow attacking speed, a mediocre overall xG. I said they would go out in the group stage. Social media laughed. On the night of June 27, 2026, Germany lost 0-2 to South Korea despite 26 shots, and were eliminated. I did not see Germany lose. I saw numbers that do not lie. But what I learned was not that I was right. What I learned is that a data argument is worth something only when it states clearly what it rests on, and what it does not know. And here is what I do not know about Vietnamese athletics. I have no training-load data from the centres. I have no recovery metrics, no injury records, no accumulated training age for any athlete. I have no 60m test times, no lactate readings. All I have are competition marks and public results. A model built on competition data always answers what happened well and why very poorly. This is where I have to be blunt: my data is incomplete, and anyone who claims otherwise is selling you something. The shoe story is the clearest example of that gap. The race for carbon-plated shoes and supercritical foam has reshaped world distance running over more than a decade, with records falling in succession in the marathon and half marathon. But I have no way to separate the equipment dividend from the athlete's ability in SEA Games data, because organisers do not collect equipment information per athlete. If I cannot separate it, I am not allowed to add it to the performance. This is the principle I have kept since my football days: never attribute to an individual something that belongs to the context. When football stopped and I switched to counting strides, I carried one belief from Hai Phong — a place with no title glamour, only spreadsheets and people working quietly. Hai Phong taught me: the star is not on the shirt, it is in the metric. In athletics that holds in a harsher way. There is no attack, no defence, no teammate screening you. There is one person in one lane and a number on the scoreboard. But for that very reason athletics data is fairer: it does not let the aura of a big collective hide an individual's dullness, nor does it let a big name be forgotten merely for lack of an audience. On the other side, I think we are missing the most important group of athletes. Counting podium faces, I noticed names that appeared exactly once and then vanished from the next edition's results. They are usually finalists without medals, athletes who set personal bests at the Games and never made the papers. In the data, that is the most important signal about the future. A system that throws up many finalists is a system that is growing. A system with a single gold medallist is a system in dependency. If I had to pick one column to track over the next four years, I would pick the number of events in which the delegation places at least two athletes in the final. That column is suspiciously low in my data. It says something few want to hear: our peak is held by a handful of individuals, not by a stream. That does not mean those individuals are unworthy. They are worth more than the numbers show, because competition marks do not measure the hours they put in. It means that if one of them is injured, changes direction, or simply ages, the gap will not fill itself. There is one counter-intuitive angle I must handle carefully. My data shows that squad depth in athletics does not correlate with gold count in the short run. A delegation can win many golds while the base is thin, and that is entirely normal within a single edition. The danger lies elsewhere: the correlation only appears when you stretch the time series to ten years. That is why I do not draw conclusions after one SEA Games, and why I say not enough data more often than I make forecasts. Numbers are a mirror. Most of the market looks into it and sees only itself. Four gold medals in four days in Phnom Penh is a fact. But it is a fact about one person, not about a system. When I close the data file and look at the column counting events with two finalists, I see an open question still intact. Over the next four years, if the number of podium faces rises while the gold count dips a little, I will call that progress. That is what I will be counting.